This project develops and evaluates a cognitive behavioral approach to the prevention of high risk sexual and drug use among Black, Hispanic and non-Hispanic White couples in south Florida who are at high risk for HIV/AIDS. Grounded in the health belief model, learning theory and stage model of behavior and behavior change, the intervention is intended to (1) facilitate increased knowledge of HIV-AIDS and perceived vulnerability, (2) reduce risky sexual and drug use behaviors, (3) increase the negotiating skills of both partners as related to the avoidance of risky behaviors and (4) increase self- efficacy and reasons for change within a culturally and gender sensitive couple context. The intervention is also intended to reduce relapses of risky behaviors. The intervention, covering eight hours over a period of four weeks, consists of four small group meetings lead by trained peer leaders, and includes HIV/AIDS knowledge. gender and cultural/social barriers to eliminating high risk behaviors and skill building in assertiveness and negotiation. Outcomes to be affected include high risk sexual behaviors (unprotected oral, anal and/or vaginal sexual intercourse and trading sex for money, drugs or favors), drug use, substance abuse that may lead to risky behaviors, HIV/AIDS knowledge, self-efficacy, assertive/negotiation skills. HIV and STD status and power and relationship satisfaction. Data will be collected through self-reports. with measures repeated at pretest, weekly during intervention sessions, posttest (at the end of the intervention) and at 3, ,6 and 9 month follow-ups. Two booster sessions are planned for 6 weeks and 18 weeks following the post-test. Varying methodological strategies are proposed to enhance reliability and validity of the self-reports and comparisons of the reports between partners. The sample will consist of ethnically balanced numbers of volunteer couples recruited from various agency sites (including public assistance agencies. STD clinics and county public health units) and their sexual partners randomly assigned to either treatment or control groups. Data analysis will employ ANOVA and structural equation models appropriate to level of measurement of the dependent variables to examine the immediate and longer-term outcomes and behavioral impacts of the proposed intervention as well as to assess the general applicability of the intervention across race/ethnic groups and gender. Both individuals and couples will be used as units of analysis. The proposed study examines 660 couples and their sexual partners using a 2 x 3 factorial design with couples randomly assigned to treatment and control groups. The intervention to be tested targets gender behaviors and the interactions between genders that occur within sexual relationships in the context of racial/ethnic and cultural norms. Results from this study will contribute to the existing. but limited data base. regarding effective strategies targeted at couples for preventing the spread of HIV infection among culturally diverse coupled a- risk for HIV/AIDS.